Public bug reported:
Apport was unable to send an automatic bug report (I am behind a
proxy). However, I will attach by hand most of what it would have sent
anyway. But first, a description of the problem:
DistroRelease: 9.04
Package and machine architecture: amd64, linux-image-2.6.28-11-generic
2.6.28-11.38
Machine: Gateway M-6862
Suspend:
------------
The machine appears to suspend fine. On wake-up, it goes right up to the
appropriate password prompt (unlock screen) then freezes there. The mouse is
responsive, but I cannot get keyboard focus to enter the password. Previously,
on this same machine, the screen would lock up completely on resume. I cannot
recall the exact kernels that behaved this way. This happens regardless of
whether I suspend from the GNOME menus or from the command line. (WAKE UP
FAILURE)
Early in the jaunty development cycle (around alpha-4: I cannot recall
exactly when), I ran with the -radeon open source driver. Suspend and
resume worked fine (after I had disabled atieventsd). I ran Intrepid to
late January, still with fglrx as my graphics driver. Suspend/resume
worked then. My memory fails me a little here: If I recall right, the
suspend/resume only started working after I manually installed a newer
version of ATI Catalyst from amd.com in January: with the catalyst in
the ubuntu repos, it did not.
Hibernate:
--------------
Hibernate has NEVER worked on this machine. Not in Intrepid, not in Jaunty. I
will describe the current symptoms, present since I moved to 9.04 alphas in
January. On hibernate, the kernel panics. I have attached a digital photo of my
screen. The panics are reproducible 10/10 (tested by repeatedly suspending and
resuming).
The following may be slightly "unusual" about my machine:
- I am running ext4. Initially, my ext4 partitions were converted from ext3.
However, while attempting to define the exact circumstances under which some
ext4 freezes occured (on other launchpad bugs), I dropped and recreated the
data partitions from scratch. So, I now have 3 "native" ext4 partitions and 1
"converted" partition (/).
- I have 2 swap partitions (side effects of evicting Windows Vista bit by bit).
Between them, they are more than twice as large as my RAM. Initially, I had a
single swap partition of 5059 MB (RAM 4GB). However, once, while mucking
through the log files, I thought I saw a message to the effect that hibernate
aborted because of inadequate swap space. I could not expand the existing swap
partition, so I created a new one.
suspend_test script
--------------------------
I ran this script. (/usr/share/checkbox/suspend_test)
- In test 1 (suspend then wake up after 20 seconds), the machine suspended but
did not wake up
- When I woke it up manually, I got the unresponsive login screen described
above
Is fglrx the culprit?
------------------------
Probably. Given the number of fglrx related complaints on Launchpad now. But,
before I condemn it, I will switch to the open source drivers and test
hibernate and suspend again (after posting this message).
I will also attach an fglrx-related kernel oops that I have consistently
gotten every time I have booted in the recent past. This is for purposes
of context only - Apport's automated bug reporting has already resulted
in a rash of this particular oops on Launchpad by now.
** Affects: linux (Ubuntu)
Importance: Undecided
Status: New
--
Hibernate AND suspend/resume failure on Gateway M-6862 laptop (Non-free kernel
module: fglrx, ATI R6xx graphics card)
https://bugs.launchpad.net/bugs/350788
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
--
ubuntu-bugs mailing list
ubuntu-bugs@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
